Investigations
Claims that circulate faster than their evidence. Each volume separates what the record supports from what got attached to it later, and says plainly which questions are still open.
er·rat·ic
/ɪˈratɪk/ noun · geology
A boulder carried far from its bedrock by ice and left standing in a landscape it has no business being in. Geologists find them by asking the obvious question: how did this get here?
Some evidence
doesn’t fit where
we found it.
Twenty-one series on secret societies, contested archaeology, vanished civilizations, institutional failure, and the histories that don't survive a close reading. Written by James D. Sutton. Deeper than a summary, more honest than the hype.
Four lines, fourteen series
Cross-section of the catalog
Civilizations
One people, one place, one volume at a time. Regional histories built from the current scholarship rather than the version that settled into the textbooks a century ago.
Retold
Ancient texts rendered as continuous narrative for readers who want the story without a critical apparatus in the way—with the manuscript problems flagged rather than smoothed over.
Institutional Failure
Investigations built from filings, transcripts, depositions, and the public record. The subjects are notorious; the method is documentary. Where the record runs out, the book says so.
Counterfactual & Speculative
What the evidence permits us to imagine. Collapse scenarios traced through their real mechanisms, and turning points followed down the road history didn't take.
How we work
-
Sources travel with the book
Every title ships with the works it was built from. The full annotated bibliography lives on this site, one page per book, with a note on what each source is good for and where it falls short.
-
Settled, contested, rejected
Each investigation marks which of its claims the evidence actually supports, which remain genuinely open, and which popular versions we're setting aside—and why.
-
No invented precision
A number appears only when a source supports it. No confident-sounding statistics manufactured to fill a sentence, and no round figures dressed up as findings.
-
Edited, not generated and shipped
Every manuscript goes through line editing against a fixed house style before it reaches a reader. Production tooling drafts; a human decides what stays.
